School is all about signaling, not skill-building • Hacker News Board

Re: School is all about signaling, not skill-building

#121

Post by ilaksh » Sat, Mar 16, 2019, 1:11 PM UTC

TAForObvReasons wrote: post #71

JMTQp8lwXL wrote: post #23

Maybe for the elite. But for middle class people like myself, I developed marketable skills, secured multiple internships, and found it easy to get a full-time job straight out of school. I had about a years worth of industry experience via internships (one for 6 months, two others 3 months each). Signaling doesn't buy you much. At least in engineering, because a technical interview will sort out whether or not you a…

> Software Engineering is a notable exception, most other fields don't quiz as part of the job interview process: soft skills and credentials or state licenses or bar exams are relied upon more heavily. Maybe signaling matters more in these fields? License and bar exams are more serious than people in technology are willing and ready to admit. They serve the role of quizzes because the "passing score" requires a base…

The first issue is that JS has evolved every few years. For example, we went from callbacks to promises and now to async/await. Or look at how modules have evolved. Or the use of const and let vs. var.

The other issue is that there are layers of other skills that also evolve every few years.

So to create a "bar exam" for a front-end web engineer, for example, you would need a completely different version every two years or so.

But then probably a bigger problem is that software engineering requires certain types of problem solving and to get a realistic idea of problem solving abilities of a person at large enough scale problems you would need a huge investment into the testing infrastructure and varied content. You would need sophisticated programs to test the applicants programs. And people to review code. And those tests would quickly go out of date.

Then the other thing is that there is so much variation in the types of languages or frameworks or tools used, it is questionable whether it is meaningful to test on some lowest-common-denominator set of tools at all, if you could find enough agreement on such a thing (e.g. Angular vs React, React vs Vue).

Re: School is all about signaling, not skill-building

#122

Post by dredmorbius » Sat, Mar 16, 2019, 1:17 PM UTC

There's a long history behind education, types of it, and the reasons, rationales, and interests behind these. One of the better synopses of these I've found is Wes Cecil's "Myths of the Modern American Mind: Education" (61 minutes, audio only):

https://www.youtube.com/watch?v=CFiiOm6fB3c

He traces the history back to Babylonia and scribal training, Plato's Academy, Aristotle's Lyceum, medieval education, scholaticism and the rise of the Seven Liberal Arts with their Trivium (grammar, logic, rhetoric) and Quadrivium (Maths, Geometry, Music (or harmony) and Astronomy), the emergence of the modern physical and social sciences, and the Prussian Educational System which is the foundation of contemporary Western education.

https://en.wikipedia.org/wiki/Liberal_arts_education

Throughout virtually all of this there's been a sharp division between intellectual and technical education. The liberal arts were distinguished from the technical arts, also called Artes Mechanicae (mechanical arts), servile arts, or vulgar arts, originally: tailoring and weaving, agriculture, architecture and masonry, warfare and hunting, trade, cooking, and blacksmithing and metalurgy.

https://en.wikipedia.org/wiki/Artes_Mechanicae

Since 1800 or so, the distinctions have largely been between pure skills, including trade, vocational, and technical skills (including STEM), professional training (law, medicine, military, religion, and general administration), and humanities or liberal arts. The relative prestige of these has shifted fairly markedly over this period, as well as the increased specialisation within academic curricula and programmes, with the University of Virginia offering eight fields of study in 1825, and Johns Hopkins University using the term "major" for the first time in 1877. The emergence of technical schools, such as M.I.T. (1861) with specialised courses of study helped drive this transformation.

Cecil's treatment is on the light side, but engaging.

John Stuart Mill's observations on the indoctrinational role of education are also interesting. As others have said, schools play multiple roles.

https://old.reddit.com/r/dredmorbius/comments/6x7u6a/on_the_...

http://www.tandfonline.com/doi/abs/10.1080/00346760110081599

Re: School is all about signaling, not skill-building

#123

Post by chongli » Sat, Mar 16, 2019, 1:17 PM UTC

bko wrote: post #116

dogma1138 wrote: post #96

Earlier quoted context omitted.

Would you go to a doctor that learned medicine form YouTube? I think the problem here is a bias due to the fact that the employment after earning their computer science degree is for many people not about the science part of it but rather about fairly mundane programming and general office work. Good luck arguing that chemistry, biology, mechanical engineering, medicine and many other fields that require specific too…

> Would you go to a doctor that learned medicine form YouTube? Depends on my ailment. Most of the time people go to the doctor they are told to get some rest and maybe prescribed something. I don't think that needs 8 years of medical school. And doctor's don't even necessarily do a good job at prescriptions either, since they tend to over-prescribe in the US

they are told to get some rest and maybe prescribed something. I don't think that needs 8 years of medical school.

What takes 8 years of medical school is knowing when not to tell the patient that. When telling the patient to go home and rest could result in severe illness or death.

Re: School is all about signaling, not skill-building

#124

Post by barry-cotter » Sat, Mar 16, 2019, 1:18 PM UTC

Godel_unicode wrote: post #37

glglwty wrote: post #11

Technologies like online education would crush the existing education system if the playground is fair. Unfortunately the latter has public funding in it and prevents the society from moving forward.

Except online education is dominated by the existing players, and they are charging the same kind of money you would pay in person. If you're getting a STEM degree especially, Coursera will charge you something like $600/credit. This is because of, wait for it, signaling. You don't actually need all that much education to be an entry level software engineer, especially outside silicon valley. But you do need those le…

Georgia Tech On Campus M.Sc. CS $40,000

Georgia Tech Online M.Sc. CS $8,000

Re: School is all about signaling, not skill-building

#125

Post by JoeAltmaier » Sat, Mar 16, 2019, 1:19 PM UTC

rjf72 wrote: post #72

hooch wrote: post #16

“If a student wants to study at Princeton, he doesn’t really need to apply or pay tuition. He can simply show up and start taking classes. As a professor, I assure you that we make near-zero effort to stop unofficial education; indeed, the rare, earnestly curious student touches our hearts. At the end of four years at Princeton, though, the guerrilla student would lack one precious thing: a diploma. The fact that alm…

One key thing this paragraph misses is the characterization of this sort of person. Imagine we have a person that would independently, without extrinsic reward or 'push', show up to classes completely of their own accord. And we must further assume that they would engage in all assignments and somehow try to regularly test their understanding - the feedback exams offer is crucial to demonstrating understanding. And w…

That's silly. Of course you have a filter for those likely to succeed. So do all institutions in the same business of the same tier. Then, you compare them on what they did with that human capital. Some could do better than others.

Just because I build a cabinet out of good, knot-free straight boards doesn't mean the cabinet is all about the boards.

Re: School is all about signaling, not skill-building

#126

Post by mikeash » Sat, Mar 16, 2019, 1:22 PM UTC

kevintb wrote: post #33

I strongly disagree. Elite colleges are all about signaling. But I don't doubt for a second that my CS degree taught me an enormous amount - some skill-building, lots of abstract thinking - that I wouldn't have been able to achieve as quickly as anywhere else.

Came here to post this. I did my undergrad at an obscure state university with relatively low admissions standards. I learned a shitload. Could I have done that on my own? In theory, yes. In practice, I’d already been doing that throughout high school and, while that gave me a good foundation with the basics for college and let me skip past the Programming 101 stuff, it never came anywhere close to the rigor and focus I got in a formal setting.

People seem to forget that the vast majority of college grads didn’t go to an Ivy League school.

Re: School is all about signaling, not skill-building

#127

Post by mikeash » Sat, Mar 16, 2019, 1:24 PM UTC

nafey wrote: post #41

kevintb wrote: post #33

I strongly disagree. Elite colleges are all about signaling. But I don't doubt for a second that my CS degree taught me an enormous amount - some skill-building, lots of abstract thinking - that I wouldn't have been able to achieve as quickly as anywhere else.

Okay but how about this: Say you actually dropped out of your degree one day before completing it. Skillwise you are almost the same but your compensation will take a drastic hit on average.

What does it say about a person who makes such a catastrophically stupid and easily avoidable decision? Will they decide to treat your big launch day like their degree, and bug out the day before never to be seen again?

Yes, this is signaling, but really useful signaling. School can provide both signaling and skills.

Re: School is all about signaling, not skill-building

#128

Post by barry-cotter » Sat, Mar 16, 2019, 1:28 PM UTC

lifeisstillgood wrote: post #49

I think the article seriously underestimated the effort to make it through a thre year degree at an elite college - let's say an MIT mechanical engineering degree - my understanding it is 80 hour weeks, with yearly inflection points where the faculty will try and eliminate the bottom 10% or so. get a first or 2.1 there and you are not signalling attendance but genuine effort. (Any alumni able to comment? ) There are…

The only élite university in the US with a significant drop out rate is Cal Tech. I’m sure there are lots of people who went to MIT who graduate with a softer degree than they expected to get going in but drop out rates at the tip top are low. At top graduate schools they birder on non-existent. Most years no one drops out of Yale Law School.

Re: School is all about signaling, not skill-building

#129

Post by barry-cotter » Sat, Mar 16, 2019, 1:35 PM UTC

NoblePublius wrote: post #8

Admissions based on standardized test scores ONLY. Make colleges co-sign loans. Fixed it.

If the university’s and students’ interests are aligned there’s no need to mandate what admissions are based on, universities will admit those they can help (most). An income share agreement á la Lambda School is clearly superior here to loans. The university is far better able to shoulder risk than an ignorant student. The university can diversify its risk across the entire student body. Individual students can’t.

Re: School is all about signaling, not skill-building

#130

Post by hopler » Sat, Mar 16, 2019, 1:37 PM UTC

barry-cotter wrote: post #115

tokyodude wrote: post #104

Earlier quoted context omitted.

Note: I am not agreeing nor disagreeing with the article. I'm only clarifying what it's trying to say The article is about the fact that school does not show that "developer ... is signaling that he is reliable and comes to work every day" The whole point of the article is there is zero correlation between people who've graduated from college and people who haven't in term of how well they do on the job. It sounds co…

> The whole point of the article is there is zero correlation between people who've graduated from college and people who haven't in term of how well they do on the job. ... among the population who have been hired to do the job. This is conditioning on the collider. The people without a degree who got in are a highly selected sample compared to all people without degrees. Anyone who used this to argue that a degree…

> Anyone who used this to argue that a degree wasn’t a strong useful signal

Reread the title of the article.

The question is whether an individual would benefit from college even if everyone's college was enrollment/graduation status was kept secret from everyone, even classmates.